Odonatologica 7 (1): 27-47 March 1,1978 Thermal adaptationsof dragonflies M.L. May Departmentof Physiology and Biophysics, University of Illinois at Urbana-Campaign, 524 Burrill Hall, Urbana, Illinois 61801, United States Received and Accepted September 26, 1977 Most Odonata probably exert some control over their body temperature. Such behaviour is favored by the high, variable temperatureof their terrestrial and habitat by the fact that dragonflies are relatively large insects and thus ex- change heat with the environment comparatively slowly. Different species may thermoregulate behaviorally by controlling the external heat load or physiol- ogically by altering the amount of heat generated by or dissipated from the thoracic muscles. The former method is characteristic of perchers, which are usually in a situation that permits considerable modulation of intercepted solar radiation, mostly by postural adjustments. There is also evidence that some species may perch more frequently in the shade when ambient tem- perature is high. Fliers, by contrast, have little opportunity to control inter- cepted solar radiation but continually generate large amounts of heat during flight. They may alter their rate of heat loss by controlling haemolymph flow from the thorax or their rate of heat gain by changing the proportion of gliding vs. flapping flight. Some species combine behavioral and physiological modes of regulation.Adaptive geographic, seasonal, and inter-habitat variation occurs in the responses of dragonflies to temperature.Voluntary avoidance of high temperature seems most closely adapted to the thermal environment. and Temperature responses thermoregulatory ability also vary with age and sex. The effects of temperature on the activity patterns of Odonata and the adaptive significance of thermoregulationhave yet to be rigorously demon- strated in most cases. INTRODUCTION Some CORBET discussed years ago (1963) adaptations to temperature of adult dragonflies and mechanisms by which they control their thoracic tem- perature (Tb). At that time, no data existed on the temperatures that dragon- flies experience in the field or on the limits of thermal tolerance. Since then, 28 M.L. May Corbet’s suppositions have been largely confirmed. Considerable information has accumulated showing that Anisoptera thermoregulate, some quite effective- ly, and that they are adapted to function at rather high Tb (MAY, 1976b, 1977). I will to of the selective forces In the present paper attempt identify some that influenced the evolution of thermoregulation in Odonata, the principal mechan- isms by which dragonflies have met the thermal challenges of their environment, which to related to and the ways in adaptations temperature are specific features of the of well out knowl- biology various species, as as point some gaps in our edge of their thermal biology. Thermoregulation is the maintenanceby an animal of body temperature rela- tively independent of ambient temperature, by means of specially-adapted be- havioral or physiological responses. Body temperature need not be held constant be uninfluenced environmental or by temperature. In fact, Tb is never absolutely of independent environmental conditions, even in very specialized thermo- like regulators man. Thermoregulation does require an active response by the animal. HEATH (1964) showed that even inanimate objects in a heterogeneous environment can have a temperature distribution that could be interpreted as indicating thermoregulation. Thus it is necessary to determine the mechanisms which controlled least show that the distribution of could by Tb is or at to Tb not occur passively. Thermoregulation is indicated if a linear regression of Tb on environmental temperature is significantly less than 1.0. As a matter of convenience I will use air in the shade in the of the insects temperature (Ta) general vicinity as a measure of effective environmental ; BAKKEN & GATES, temperature (Te In cases T and T are not since T is also influenced 1975). many a e equal, e by radiant and heat T is conservative estimate evaporative exchange. Generally a a of T since the latter is more variable. T for T thermo- e , usually By substituting a e that be but it is that insect regulation actually occurs may obscured, unlikely an would appear to thermoregulate when it does not. All thermoregulatory mechanisms require a source of energy plus some means of controlling rates of heat gain and/or loss. Heat is exchanged with the environment via of from the (1) evaporation water respiratory or body surfaces, of radiant with other from (2) exchange energy objects, (3) conduction to or other objects, and (4) convection to the surrounding air. In all but the first case, the rate of exchange depends on the instantaneous difference between the body and its surroundings. Thermoregulating animals in general, and dragonflies in particular, may be classified as ectotherms or endotherms depending on the of source heat used in thermoregulation. Ectotherms require an external heat source, such as the sun; the latter special case is known as heliothermy. Endo- therms depend primarily on heat-producing metabolic processes. Dragonflies may use both sources simultaneously or they may switch between endothermy and ectothermy from time to time. Thermal adaptations of dragonflies 29 The size of an animal greatly influences its ability to thermoregulate and its small thermoregulatory strategy. In objects the ratio of surface area to mass is than of the greater in large objects same shape. As a result, heat exchange with animals the environment is more rapid in small and their Tb is more closely to T In addition, convectionbecomes more relative to other coupled a. important modes of heat exchange (PORTER & GATES, 1969). Some of the consequences for dragonfly thermoregulation will be explored below. METHODS Techniques for measuring Tb in the field, maximum voluntarily tolerated temperature (MVT) and heat torpor (HT), endothermic warm up rates, effects of circulation on heat transfer, and cooling constants (K) have been described (MAY, 1976b). Maximum voluntarily tolerated temperature is the temperature at which dragonflies heated with a lamp in the laboratory act to avoid further increase in Tb- Heat torpor is the point at which paralysis occurs as a result of overheating. All new experiments on MVT and HT were performed the same that the animals The last defined day were captured. parameter, K, is a constant - Newton’s law of = T where t is time in minutes. by cooling, dTb/dt K(Tb a), This relationship accurately describes passive cooling in dragonflies under most circumstances (MAY, 1976b, 1976c). The cooling constant can be converted to thermal conductance, a measure of the ease of heat transfer, by multiplying by the heat of cal 1 specific insect tissue [0.8 (g^C)" ; KROGH & ZEUTHEN, 1941]. Data on variation of the thermal environment were obtained with a BAT-4 portable thermocouple thermometerand copper-constantanthermocouples. One thermocouple junction was implanted in the thorax of a freshly killed male Erythemis simplicicollis. The dragonfly was mounted in a lifelike posture on a of bark and second the bark the chip a thermocouple was taped to so junction was about 2 cm from the dragonfly and 1 cm above the substrate. The assembly was placed on an exposed log at the shore of a lake. Other ambienttemperatures were taken with a thermocouple mounted in a hypodermic needle. All thermo- calibrated accurate couples were against an mercury thermometerbefore use. The effects of the of postural adjustments on area body surface exposed to sun was determined in Pachydiplax longipennis. I tethered individuals, with a short of wire in the of length implanted thorax, on a piece tracing paper taped the of box. The entire to top a Plexiglass apparatus was placed .outside on a 45° hot, sunny day. A mirror was placed at an angle of approximately beneath the insect and the reflection of the dragonfly’s shadow was photographed. As they warmed to levels exceeding the MVT, some individuals assumed the obelisk posture (CORBET, 1963; MAY, 1976b) that is a characteristic heat avoidance this The difference of posture in species. in the area the shadows of individuals oriented horizontally and the same individuals in the obelisk represents the 30 M.L. May proportional reduction in effective area exposed to solar radiation. The approxi- mate maximum and minimum cross-sectional areas of the tagmata of Pachy- diplax were calculated from external measurements made with an ocular micro- meter. of determined of Wing areas various species were by making pencil rubbings excised wings in paper of known weight per unit area. The outlines obtained were cut out and weighed. Wing loading was calculated by dividing the body of each mass specimen by its wing area. The effect of temperature on the tendency to glide rather than continuously examined Tramea carolina. On several flap the wings was in sunny days near Gainesville, Florida, I observed flying individuals far from breeding sites. I followed each individual as closely as possible with binoculars. Every 5 or 10 seconds I noted whether the wings were moving at the instantof observation. RESULTS AND DISCUSSION SELECTION PRESSURES Three major aspects of dragonfly biology seem critical in the evolution of Adult terrestrial thermoregulation. dragonflies are and diurnal, they are power- their aerial ful fliers and depend on ability during most activities, and they are comparatively large insects. These are all characteristics that favor thermo- regulation. The terrestrial and diurnal habitat presents both the necessity and the opportunity for thermoregulation. Because of the high density and specific heat of water, the aquatic environment from which the ancestors of insects arose was thermally more stable than the environment into which they spread. the animal At the same time, heat exchange between and the surrounding medium much reduced Thus terrestrial is in air as compared to water. organisms can maintain a sizable difference between and T provided they are exposed a, to some heat source such as sunlight or metabolic heat, much more easily than level environmental aquatic organisms.